GART TLB error generic level generic

Clydesdale · 08-13-2007, 06:24 PM

Posted this in hardware too, we really don't know if this is a software issue or hardware. Think it's bogus but...

Having an issue with several LS20 blades, AMD, Running Red Hat Linux 3.2.3-53. While investigating other issues we found this error on a number of the machines. We do not think it is memory related but are not certain. Bios are updated to current levels:

Aug 8 09:07:42 xtky8205pap kernel: CPU 1: Silent Northbridge MCE
Aug 8 09:07:42 xtky8205pap kernel: Northbridge status a6000001:0005001b
Aug 8 09:07:42 xtky8205pap kernel: Error gart error
Aug 8 09:07:42 xtky8205pap kernel: GART TLB error generic level generic
Aug 8 09:07:42 xtky8205pap kernel: err cpu1
Aug 8 09:07:42 xtky8205pap kernel: processor context corrupt
Aug 8 09:07:42 xtky8205pap kernel: error uncorrected
Aug 8 09:07:42 xtky8205pap kernel: previous error lost
Aug 8 09:07:42 xtky8205pap kernel: NB error address 0000000037ff0008
Aug 9 11:17:49 xtky8205pap kernel: nfs_safe_remove: Engine/.#ds_updating busy, d_count=2
Aug 10 03:24:52 xtky8205pap kernel: nfs_safe_remove: Engine/.#ds_updating busy, d_count=2
Aug 10 10:38:16 xtky8205pap kernel: CPU 1: Silent Northbridge MCE
Aug 10 10:38:16 xtky8205pap kernel: Northbridge status a6000002:0005001b
Aug 10 10:38:16 xtky8205pap kernel: Error gart error
Aug 10 10:38:16 xtky8205pap kernel: GART TLB error generic level generic
Aug 10 10:38:16 xtky8205pap kernel: err cpu0
Aug 10 10:38:16 xtky8205pap kernel: processor context corrupt
Aug 10 10:38:16 xtky8205pap kernel: error uncorrected
Aug 10 10:38:16 xtky8205pap kernel: previous error lost
Aug 10 10:38:16 xtky8205pap kernel: NB error address 0000000037ff0010

The blades do not show any outwards signs of having any kind of stability trouble etc. Any help would be greatly appreciated.

CD

unSpawn · 08-13-2007, 06:47 PM

Please post your thread in only one forum. Posting a single thread in the most relevant forum will make it easier for members to help you and will keep the discussion in one place. This thread should be closed because it is a duplicate of http://www.linuxquestions.org/questi...d.php?t=576863.